What John Oliver Got Right – And Wrong – About Subprime Autolending

The concept of sub-prime borrowing is a bit tricky.

In general, the term  “sub-prime” refers to borrowers with lower than average FICO scores and less than stellar payment history.

What exactly “less than stellar” means is up to some interpretation as it can actually corral together a wide variety of borrower profiles. On one end of the spectrum, it can describe a range of borrower behaviors that includes those with a long and consistent record of not paying bills on time or at all –  and on the other end of the spectrum, it includes those who went through a temporary run of bad financial luck that set off a lot of late payments and defaults.

This is why there are a variety of numerical ways sub-prime borrowers are categorized. By some measures a sub-prime FICO score is any score under a 600; by other measures under 580 is a sub-prime score. Some scales differentiate between sub-prime borrowers with scores below 620 and deep subprime borrowers with scores below 500. For a brief window after the credit crisis in 2008-2010, any credit score under 700 was basically considered subprime.

However one chooses to break it out, it is clear that by any measure, lending to sub-prime borrowers is on the rise, and enough so that it is catching a lot of people’s attention.

Last week the New York Fed released data that indicates household debt is sharply on the rise – and the borrowing among subprime borrowers has been particularly active. The NY Fed also noted that credit card spending and the prolonged surge in auto-lending were particularly driving the new debt boom.

That release appeared at roughly the same time that JP Morgan Chase and Company CEO Jamie Dimon warned investors that conditions in auto-lending as an area are such that his big bank is treading very carefully.  Singling out the auto-lending spike, Dimon noted, “Someone is going to get hurt,” he said. “It won’t be us.”

Jamie Dimon never fails to be colorful in his remarks.

But on this particular issue, he didn’t actually win the award for most colorful remarks on the subject – that award surely goes to British comedian and infotainment champ John Oliver. Oliver spent much of yesterday receiving his weekly round of applause from Slate, Salon and the Huffington Post for his “epic takedown” of auto lending for subprime buyers  in all of its borderline predatory glory.

And, for a comedian, Oliver got a surprising amount of stuff right – and did reveal some legitimately ugly stuff. In fact – Oliver might just be righter than he knows – depending on which way default data goes across the board.

Oliver also skips over some things and doesn’t quite tell the whole story on others – he is probably right to warn viewers that he’s an entertainer not a journalist.

But we are, so we thought we’d give you the quick primer on what John Oliver got right, and what still needs clarification about sub-prime auto lending.

The Case Against Subprime Auto Lending

For a man who bills himself as a professional comedian – John Oliver makes some fairly insightful and disturbing points about the wide-world of sub-prime auto lending.

For those who can’t watch, Oliver says that in a car dependent society – the costs of getting a car for sub-prime buyers can run extremely high.  Noting that as of 2016 about quarter of all car loans qualify as subprime and that there seems to be a proliferation of auto businesses built entirely on selling and lending to subprime buyers, Oliver points to the variety of shady tactics that are observably hovering around the industry.

And shady is barely putting a fine enough point on it.  Consumer horror stories abound throughout Oliver’s report including a consumer who, through gross upcharging and a 24 percent interest rate, faced paying $17,000 for a $3,000 car over 6 years. Other highlights included the woman who’s dealership repossessed her car with her baby inside or the “demon car” that was bought, and repossessed and rebought over 10 times in in 6 years.

These horror stories all issued from so-called “buy here, pay here dealerships,” and Oliver reported (on the basis of NABD data) that around 33 percent of these loans default, and that the majority of loans that do default will do so within the first seven months of the loans. And Buy-Here, Pay-Here was even good enough to cough up a villain straight out of central casting in the form of Ken Shilson, the President of the National Alliance of Buy-Here, Pay-Here Dealers.

Caught on hidden cameras speaking at a conference in Las Vegas, Shilson gave some rather world-class and insensitive remarks on the wonderful investment opportunity his industry’s irresponsible lending habits presents when combined with other people’s life problems:

“These people have negative equity all the way through and they’re not going to be able to hold their life together for the next seven years. You know the customer.  You know how to collect. You got the legal expertise. What’s holding you back?”

Oliver also noted that these issues were particularly severe among buy-here, pay-here dealers and that these arrangements were coming out of fashion somewhat as mainstream lenders across the board – as well as specialty startups – are heavily pushing into the market and simply out-competing their less reputable cousins.  More focused on “upper echelon” subprime buyers with credit scores +/- 20 points of 600, banks like Santander and sub-prime focused start-ups like Skopos and Exeter offer high cost – but not usurious – rates to subprime buyers.

However, he also notes that the subprime lending boom bears some structural similarities to the subprime lending boom in housing that led to the financial crisis – particularly the habit of packaging subprime loans into bonds that can then be purchased.

This is what Jamie Dimon was referring to when he noted “someone is going to get hurt,” noting that the rapid uptick in enthusiasm is leading to the kind of exuberance that causes too much credit extended to the wrong borrowers – leading to defaults.  If too many loans default, the bonds become worthless – and if enough money goes up in smoke – well, we all saw what the Crisis looked like when it was mortgages in 2008.

But there are many, many reasons (that Oliver sort of glossed over) that auto loans aren’t the new Autocrisis.

What John Oliver Got Less Right  

The main issue – and one Oliver points out (as does Jamie Dimon and many, many others) is one of scale.  The auto lending market has soared in value over the last six years. In the last year alone the market has grown over 11 percent and is now worth over a trillion dollars.  That, noted Dimon, makes it far less a risk of being a “a systemic issue” because it is a smaller part of the system. The total auto-loan market is about 12 percent the size of the $8.4 trillion mortgage market as of the first quarter, according the Federal Reserve Bank of New York.

And the vast majority of those loans are held by prime borrowers who face none of these issues – and who weren’t purchasing cars during the recession.

And as we noted, subprime is a more complicated designation that John Oliver’s commentary let on.  Alternative scoring model, VantageScore, shows that a lot of what’s taken for granted in traditional credit scoring isn’t as good a predictor of who will and will not pay a loan as people assume.  SVP of Research, Product Management and Analytics Sarah Davies noted that swings in a credit score as much as 20 points – enough to put customers into and out of prime borrower status – can and do routinely occur in credit scoring and are basically meaningless to answering the question of whether or not a customer is a good credit risk.

VantageScore also notes that a score is only a number – and lenders concerned about managing defaults needs to be more wired into why the number is what it is – instead of merely checking it off.

“With the understanding that a credit score is simply a proxy for risk, the fact that a consumer’s credit score changes or doesn’t change is only meaningful in the context of understanding the actual risk estimate associated with the credit score at any given point in time,” a recent VantageScore white paper notes.

Arguably, if subprime borrowers ever want to move up a category, someone is going to have to lend to them.  And although subprime borrowers can be a risk factor for additional default – and waves of defaults are very destabilizing – Vantage’s data indicates that merely being subprime is not necessarily indicative that borrowers are going to start defaulting in waves.

And so far the data on defaults is mixed.  On the whole, default rates in auto lending have gone up – but only very slightly, to 2.1 percent.  Lenders that specialize in – and actively recruit – deep subprime buyers have seen bigger spikes in the 12 percent range, but outside those specialty niches the figures haven’t been dramatic.  Mainstream lenders underwriting subprime auto-buyers are seeing some increase in defaults, but not outside modeled predictions for defaults given the borrowers are subprime – an increase default rate is built into the model.

Which is the sort of essential issue with John Oliver’s story. Subprime buyers are charged higher interest rates because they tend to default – not because lenders enjoy predating on poor people.  But low income buyers with low credit scores tend are statistically more risky – hence the lender charging more interest to protect their investment in the loan.

There are certainly lenders who take advantage of that – and we met several of them in this report.  They are clearly well beyond protecting an investment and have moved to straight profiteering.

But it bears noting how the report began – the story of a woman who commutes two hours to get to work on public transport, when driving would take her less than ten minutes. It’s hard to decide if that woman we met (about whose car loan status we never learn) at the beginning of the story is paying more commuting 14 hours a week than if she had to take a high interest loan to get a car.

There aren’t easy answers.  But it seems that if any firm that is not a charity lends to high risk borrowers and prices for that risk is automatically labeled a suspicious bad guy deserving of an epic take down – well, it’s hard to imagine wanting to lend to anyone with a less than stellar credit score.

But that probably doesn’t much help subprime buyers either.  Especially if they need a car.